Drunk-driving fatalities have declined significantly over the past 25 years; however, many more victims are surviving their injuries, and it is largely unknown if similar declines in alcohol involvement have occurred among surviving trauma patients, especially for non-vehicular injuries. Nor is it known the extent to which elevated Blood Alcohol Concentration (BAC) on admission is a marker for subsequent adverse outcomes. The long-term goal of the proposed study is to improve the understanding of alcohol's role in injury causation and use this information to better target alcohol treatment services in trauma centers to reduce the number of deaths from recurrent trauma and alcohol-related disease. The objective of this application is to develop a comprehensive toxicology database on alcohol involvement in non-fatal injuries, spanning 1983 to the present, to use this data to evaluate trends in alcohol involvement in non-fatal injuries over time, and to determine how an elevated BAC on admission relates to subsequent mortality risk (especially how this risk varies by BAC level). The extended toxicology database will be used to conduct a retrospective cohort study that will link discharged cases to the National Death Index to determine if and how former trauma center patients die anywhere in the United States. Survival analysis will be used to evaluate the extent to which elevated BACs on admission predict subsequent mortality from injury or alcohol-related disease. The underlying hypotheses are that: 1) declines in alcohol involvement for traffic crashes have been largely the result of prevention efforts directed at reducing drunk driving, and that there has been less change in alcohol involvement in association with other injuries; and 2) persons with very high BAC levels (>150 mg/dL) are at much greater risk of subsequent death from repeat injuries and alcohol-related diseases than patients with lower or negative BACs. Guided by strong preliminary data, these hypotheses will be tested by pursuing three specific aims: 1) to develop a comprehensive toxicology data system spanning 25 years that will be used to document alcohol involvement among all injured patients admitted to the busiest Level 1 trauma center in the US; 2) to evaluate trends in alcohol involvement in non-fatal injuries over time and determine the extent to which the well-described trends in fatalities from drunk driving have occurred for non-fatal injuries, including motor-vehicle crash admissions and other modes of injury, both unintentional (e.g., falls) and intentional (e.g., assault and suicide); and 3) to determine, among those discharged alive from the hospital, the extent to which BAC+ status on admission predicts subsequent mortality from injuries and alcohol-related diseases. The proposed work is innovative because it will link unique longitudinal data on alcohol consumption by trauma center patients with a national death register to identify patients who die after discharge. The proposed research is significant because it will improve the monitoring and targeting of alcohol prevention and treatment programs designed for groups most at risk of alcohol-related injury or diseases.